Coarse-to-fine Categorization of Visual Scenes in Scene-selective Cortex

Abstract

Neurophysiological, behavioral, and computational data indicate that visual analysis may start with the parallel extraction of different elementary attributes at different spatial frequencies and follows a predominantly coarse-to-fine (CtF) processing sequence (low spatial frequencies [LSF] are extracted first, followed by high spatial frequencies [HSF]). Evidence for CtF processing within scene-selective cortical regions is, however, still lacking. In the present fMRI study, we tested whether such processing occurs in three scene-selective cortical regions: the parahippocampal place area (PPA), the retrosplenial cortex, and the occipital place area. Fourteen participants were subjected to functional scans during which they performed a categorization task of indoor versus outdoor scenes using dynamic scene stimuli. Dynamic scenes were composed of six filtered images of the same scene, from LSF to HSF or from HSF to LSF, allowing us to mimic a CtF or the reverse fine-to-coarse (FtC) sequence. Results showed that only the PPA was more activated for CtF than FtC sequences. Equivalent activations were observed for both sequences in the retrosplenial cortex and occipital place area. This study suggests for the first time that CtF sequence processing constitutes the predominant strategy for scene categorization in the PPA.

Effective connectivity in the neural network underlying coarse-to-fine categorization of visual scenes. A dynamic causal modeling study

“…Different analyses of the present data set were reported earlier (Musel et al, 2014). 1 Data have been fully reanalyzed in order to investigate the effective connectivity between the regions underlying coarse-to-fine processing using dynamic causal modeling.…”

“…Response accuracy (ACC) and reaction times (RT, in milliseconds) were recorded. 1 Different analyses of the present data set were reported earlier in order to investigate the spatial frequency processing in scene-selective regions (Musel et al, 2014). In fact, all participants endorsed a dynamic scene experiment followed by a separate functional localizer in order to localize scene selective regions, and a region of interest (ROI) analysis was performed.…”

“…We used a backward mask, built with 1/f noise, to prevent retinal persistence of the scene. It should be noted that stimuli and task were identical to that of Musel et al (2014).…”

“…Results of the localizer experiment and of the ROI analysis were not reported here since they took place in a different theoretical context (i.e. spatial frequency processing within category-specific regions) and were not relevant to the present goals, but can be found in Musel et al (2014). Data have been fully reanalyzed in order to investigate the effective connectivity between the regions underlying coarse-to-fine processing using dynamic causal modeling, in the theoretical context of proactive models of visual perception.…”

“…It has already been established that the fusiform and parahippocampal gyri are involved in face, object and scene recognition (Epstein, Harris, Stanley, & Kanwisher, 1999;Epstein & Kanwisher, 1998;Grill-Spector, 2003;Haxby et al, 2001;Tanaka, 1996). These regions were also recently shown to be sensitive to low-level properties of stimuli such as spatial frequencies (Goffaux et al, 2010;Kauffmann, Ramanoël, Guyader, Chauvin, & Peyrin, 2015;Musel et al, 2014;Rajimehr, Devaney, Bilenko, Young, & Tootell, 2011). These activations are therefore consistent with Bar and colleagues' model of visual perception (Bar, 2003(Bar, , 2007Bar et al, 2006) which postulate that visual analysis begins with the parallel extraction of different elementary attributes at different spatial frequencies, and that, at neurobiological level, LSF content of an image is rapidly projected through magnocellular pathways from the occipital cortex to the orbitofrontal cortex, where it activates plausible interpretations of the visual input.…”